words2graph

words2graph is a tool that processes PDF documents, extracts named entities and relationships, and visualizes them as a knowledge graph.

Installation

1. Clone the repository:

   git clone https://github.com/tomsanbear/words2graph.git
   cd words2graph

2. Run the setup script:

   ./deps.sh

   This script will:

   • Install Homebrew (if not already installed)
   • Install pyenv
   • Install huggingface-cli
   • Install Ollama and pull the llama3.1:latest model
   • Set up a Python virtual environment and install required packages

   Note: The script currently supports macOS. For other platforms, you may need to install dependencies manually.

3. Activate the virtual environment:

   source env/bin/activate

Getting Started

1. Ensure you have a PDF file you want to process.

2. Load the PDF into the database:

   python main.py load path/to/your/document.pdf

   This command will process the PDF, extract text, identify entities and relationships, and store them in the database.

3. Visualize the knowledge graph:

   python main.py visualize

   This will generate a file named knowledge_graph.html in your current directory.

4. Open knowledge_graph.html in a web browser to view the interactive knowledge graph.

Usage

The main script provides two primary commands:

• load: Process a PDF and store its content in the database.

  python main.py load <path_to_pdf>

• visualize: Generate a visualization of the knowledge graph.

  python main.py visualize

Notes

• The graph visualization shows entities as nodes and relationships as directed edges.
• Edge thickness represents the number of connections between entities.
• Hover over edges to see detailed information about the relationships.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License